It isn't so much what AI is or isn't. It's more about what can be done with it and we have already seen that a lot can be done for far less effort than was previously possible. [snip] These kinds of models are everywhere and of course include image/language processing/generation.
I know little about computer programming and even less about generative AI. Yet, the hype about AI poses it as a form of intelligence. It is not. Instead, it appears to be a copying machine specifically designed to circumvent authorship. The second part is not accidental, it is the whole point. With AI, these companies can respond to user queries by locating the answers in works produced by humans and then paraphrasing them, to avoid giving credit or paying royalties to the actual authors. Until recently, all of human culture — that is, all images, all music, all of the built environment, all technology, and all written text — was produced by individuals or teams of people. Now, we will be, ever more, confronted by culture that is produced by sophisticated plagiarising machines.
There is a potential legal issue surrounding some types of data sourcing.
'Hype' is on some ends of some AI tools but AI is far, far more than hype and the results are already here for all to see. And it's still very early days.
I provided a couple of links to just two solutions that are not connected with hype or dubious data sourcing.
Those were two Pangu model solutions from Huawei. There are currently more than 300 (many aimed at industry, health, science etc).
It's GOD Network is 'learning' on-the-go from drive data collected by cars running its systems.
If you add in hundreds more solutions from the likes of Google, Microsoft, Amazon, Tesla, etc and hundreds more from companies specialising in niche areas you quickly realise that hype is just a necessary part of some aspects. A lot of the consumer facing options need to be 'sold' as they are potential revenue streams.
It isn't applicable to AI everywhere and 'AI' itself is simply an umbrella term for many things.
Comments
'Hype' is on some ends of some AI tools but AI is far, far more than hype and the results are already here for all to see. And it's still very early days.
I provided a couple of links to just two solutions that are not connected with hype or dubious data sourcing.
Those were two Pangu model solutions from Huawei. There are currently more than 300 (many aimed at industry, health, science etc).
It's GOD Network is 'learning' on-the-go from drive data collected by cars running its systems.
If you add in hundreds more solutions from the likes of Google, Microsoft, Amazon, Tesla, etc and hundreds more from companies specialising in niche areas you quickly realise that hype is just a necessary part of some aspects. A lot of the consumer facing options need to be 'sold' as they are potential revenue streams.
It isn't applicable to AI everywhere and 'AI' itself is simply an umbrella term for many things.